EP3973255A1

EP3973255A1 - Method and apparatus for controlling/regulating a rotary drive of a work unit of a track laying machine

Info

Publication number: EP3973255A1
Application number: EP20725094.5A
Authority: EP
Inventors: Markus QUITTNER; Harald Daxberger; Reinhard BÖCK
Original assignee: Plasser und Theurer Export Von Bahnbaumaschinen GmbH
Current assignee: Plasser und Theurer Export Von Bahnbaumaschinen GmbH
Priority date: 2019-05-23
Filing date: 2020-04-23
Publication date: 2022-03-30
Anticipated expiration: 2040-04-23
Also published as: CA3135456A1; CN113874687B; EP3973255B1; EA202100237A1; AU2020279435A1; JP2022534700A; KR20220012835A; WO2020233933A1; AT522652A1; US20220235521A1; JP7485699B2; CN113874687A; EP3973255C0

Abstract

The invention relates to a method for controlling/regulating a rotary drive (13) of a work unit (4) of a track laying machine (1), in which a measurement variable (X) with an approximately periodic history function derived from a rotation of the drive (13) is detected by means of a sensor (19), a frequency (f) or period duration (T) of the history function being determined by means of an evaluation unit (21) and the frequency (f) or period duration (T) being compared with a target variable for predetermining an actuation signal. In this case a series of time-discrete measurement values (x_i) is formed for the measurement variable (X), an autocorrelation of these measurement values (x_i) for determining the frequency (f) or period duration (T) being carried out by means of a computer unit (22). Thus, in contrast to a conventional method with zero crossing detection, an accurate detection of frequency changes is also possible between two zero crossings.

Description

description

Method and device for controlling / regulating a rotary drive of a

Working unit of a track construction machine

Field of technology

The invention relates to a method for controlling / regulating a rotary drive of a working unit of a track construction machine, whereby a measured variable derived from a rotation of the drive with an approximately periodic course function is detected by means of a sensor, with a frequency or period of the

Course function is determined and the frequency or period is compared with a setpoint for specifying a control signal. The invention also relates to a device for performing the method.

State of the art

From the publication Auer F. et al .: Fligh-Tech tamping units for

sustainable track improvement, El-Eisenbahningenieur, November 2015, pages 18-22, a method for speed control of a rotary drive of a tamping unit is known. An eccentric shaft is driven to generate a vibration, which is transmitted to the tamping ax via auxiliary drives.

[03] This gives the possibility to use the

Change the vibration frequency in a targeted manner. A higher frequency (42-50 Hz) is specified when the tamping tines are immersed in a ballast bed. When ordering the tamping tines, the optimal frequency is 35 Hz. When raised, a reduced idle frequency (approx. 28 Hz) is set at which the unit runs smoothly.

In a tamping machine with speed control, there is also one

Phase stabilization known. The speeds of all vibration generators are synchronized while idling and a respective phase shift of the rotating drives is set in such a way that vibration overlap is minimized.

[05] Rotary vibration drives are also used in other work units of track construction machines. For example, WO 2008/009314 A1 discloses a so-called track stabilizer in which

Stabilization units with rotating imbalances are made to vibrate. Two synchronized stabilization units are operated with an adjustable oscillation frequency.

[06] From Hauke R. et al .: Ballast cleaning machines - an overview, El-

Special track construction machines and devices, May 2016, pages 30-35, ballast cleaning machines with various screening systems are also known. Rotary drives with adjustable vibration frequency are also used.

Summary of the invention

The invention is based on the object of providing a method of the above

mentioned type to the effect that an exact determination of the frequency or period duration can be carried out with a rapid change detection. It is also an object of the invention to provide a

specify corresponding device for performing the improved method.

[08] According to the invention, these objects are achieved by the features of the independent claims 1 and 11. Advantageous further developments of the invention emerge from the dependent claims.

[09] A series of time-discrete measured values is formed for the measured variable and an autocorrelation of these measured values is carried out by means of a computer unit to determine the frequency or period duration. Compared to a conventional method with zero crossing detection, this means that even between two zero crossings there is an exact detection of

Frequency changes possible. Function values of an autocorrelation function can be determined at any time with the recorded time-discrete measured values. The result of a corresponding function calculation are function values over a time axis. On the time axis there is a time span between zero and a first occurring maximum is the period of the

History function of the measured variable. This means that the frequency can be re-determined immediately with each newly recorded measured value.

In a further development of the invention, a further measured variable derived from a rotation of a further drive and with an approximately periodic course function is recorded by means of a further sensor, a further series of time-discrete measured values being formed for the further measured variable and a cross-correlation of the Measured values of both measured quantities to determine a phase shift is carried out. Due to the cross-correlation that can be carried out at any time, a deviation of the specified phase shifts can be recognized immediately. This means that several rotary drives are precisely synchronized

(Phase stabilization) ensured.

[11] A cycle time is advantageously specified for forming the time-discrete measured values, the cycle time determining an evaluation period. In this way an evaluation of the

Frequency or period and possibly the phase shift carried out. The accuracy of the process increases as the cycle time decreases.

[12] Another improvement provides that with each new measured value an iterative calculation of function values of a correlation function takes place by adding a constant number of measured value products. This limits the computational effort and creates the possibility for further simplifications.

[13] It is particularly useful for a current calculation of a

Function value from the sum of the measured value products of the

the previous calculation to subtract the measured value product with the oldest measured value and to add a new measured value product to the current measured value. In this way, only a few arithmetic operations are necessary to update the function values of the correlation function. Due to the low demands on computing power, the

Calculations can be carried out almost in real time with an inexpensive and space-saving computer unit. [14] Another improvement provides that before calculating

Function values of a correlation function interpolated measured values are determined. By interpolating additional measured values between the

given subsequent values, the position of the maximum of the correlation function can be determined more precisely. In this way, the frequency or period duration can be determined more precisely.

[15] The quality of the frequency determination is also increased if the

Measured values before a calculation of functional values of a

Correlation function can be filtered. For example, a

Measurement signal processing by means of a so-called Infinite Impulse Response Filter (IIR bandpass filter 4th order). A high pass eliminates a direct component of the signal, while a low pass attenuates and removes high-frequency interference in the signal. IIR filters also have the advantage that, in contrast to other digital filter types (e.g. FIR filters), they have significantly fewer arithmetic operations. This property is of great use here because of the demands on the computing power of the

Computing unit remain limited.

The method is also improved in that interpolated function values are determined before a maximum of the correlation function is determined. The interpolation is only sensibly interpolated in the area around the extreme values in order to be able to determine the position of the extreme values more precisely. The increased accuracy is thereby with little

Computational effort achievable.

[17] In an advantageous embodiment of the invention, a

Distance between a sensitive element of the sensor and an eccentric shaft driven by the drive is detected. Because of the error tolerance when determining the frequency using the correlation function, there is no

Need for precise assembly of the sensor or one

Calibration process. If the drive has an eccentric shaft

ferromagnetic material, an inductive distance sensor can be used without additional adjustments to the rotating parts.

[18] An alternative variant provides that a measured variable based on the

Sensor-acting field strength of one set in rotation with the drive Magnetic field is detected. A driven shaft is magnetized or a magnet is added to generate a rotating magnetic field. The additional space required is also when a

Magnets low. A suitable sensor near the shaft is used to record the field strength that changes during a shaft revolution.

The inventive device for performing one of the

The method described comprises a sensor for detecting a measured variable derived from a rotation of a drive with an approximately periodic profile function, an evaluation device for determining a frequency or a period duration of the profile function and an assembly for controlled control of the drive. In this case, time-discrete measured values of the measured variable are fed to a computer unit, an algorithm being set up in the computer unit to carry out an autocorrelation of these measured values to determine the frequency or period duration. This means that an exact frequency determination can be carried out immediately with each newly recorded measured value.

[20] In an improved device, a further sensor is arranged for detecting a further measured variable derived from a rotation of a further drive, the computer unit being supplied with time-discrete measured values of the further measured variable and an algorithm in the computer unit for performing a cross-correlation of the measured values of both measured variables is set up to determine a phase shift. This provides a simple arrangement for phase stabilization.

To implement the device with few and compact components, it makes sense if the computer unit is a microprocessor. Optimized algorithms and an efficient implementation of the signal processing path are used in order to adapt the computing power to the properties of the microprocessor, including the limited working memory.

The computer unit is advantageously with a

Communication interface arranged in a first assembly, with a dedicated assembly for the drive, which includes power electronics, a controller unit and a communication interface with the first assembly is coupled. The separate assemblies allow the device to be easily scaled. That way is the

Computer unit can be used for frequency or period duration determination of several drives.

For efficient detection of the measured variable, it is advantageous if the sensor includes a sensitive element for capacitive, inductive or magnetic detection of the measured variable. There are no special requirements for mounting the sensor because neither the offset nor the exact value of the amplitude of the sensor signal are relevant for further processing.

Brief description of the drawings

[24] The invention is explained below by way of example with reference to the accompanying figures. It show in schematic

Presentation:

Fig. 1 track construction machine with tamping unit

Fig. 2 track construction machine with stabilization unit

Fig. 3 tamping unit in a side view

Fig. 4 Cross section through track with tamping machine

Fig. 5 Sensor with distance measurement

Fig. 6 Sensor with field strength measurement

Fig. 7 measured values

Fig. 8 correlation function

9 Formation of function terms

Fig. 10 system illustration

11 System illustration with assemblies

12 Scheme of signal processing

Description of the embodiments

The track construction machine 1 shown in FIG. 1 is a tamping machine and comprises a machine frame 3 that can be moved on rail bogies 2. A tamping unit is arranged on the machine frame 3 as a working unit 4. The tamping machine is used to process a track 5, on which rails 7 fastened to sleepers 6 are supported in a ballast bed 8 are. During the tamping process, the track grid formed from sleepers 6 and rails 7 is lifted into a desired position by means of a lifting / straightening unit 9 and a measuring system 10 and, if necessary, shifted laterally. To fix this position, vibrating tamping tools 11 of the working unit 4 are immersed in the ballast bed 8. The submerged tamping tools 11 are placed next to one another and in the process compact ballast below the raised sleepers 6.

To generate the vibration, the tamping tools 11 are equipped with a

Vibration generator 12 coupled. The vibration generator 12 comprises a rotary drive 13 which drives an eccentric shaft 14. Additional drives 15 are mounted on the eccentric shaft 14. When rotating

Eccentric shaft 14 produces the desired eccentricity

Vibration amplitude.

After a tamping process, track 5 is usually stabilized to

Anticipate subsidence. The one shown in FIG. 2 is used for this purpose

Track construction machine 1. This is a dynamic one

Track stabilizer (DGS) with two stabilization units as

Working units 4. Each stabilization unit comprises one

Vibration generator 12 with rotating imbalances, which are driven by means of a rotary drive 13. The active one

Vibration generator 12 causes the stabilization unit to vibrate transversely to the longitudinal direction of the track. That clutches

Stabilization unit with roller tongs 16 the rails 7 of the track 5, whereby the vibration is transmitted to the track grid. This causes the track grid to vibrate into the ballast bed 8.

[28] Both in the case of tamping and stabilizing, as well as in other vibration generators 12 used in track construction, the vibration generated should correspond to various specifications. For example, a vibration frequency of 35 Hz is specified for optimal ballast compaction. For an immersion process of the tamping tools 11, a higher frequency f of approximately 45 Hz is desired in order to lower the immersion resistance.

Outside the ballast bed 8, the frequency f should be lower in order to reduce the noise pollution. [29] Further requirements are explained with reference to the working unit 4 in FIGS. 3 and 4. The working unit 4 comprises four tamping units 17, which can be lowered into the ballast bed 8 separately from one another by means of a height adjustment drive. There are 17 for each stuffing unit

Opposite tamping tools 11 are coupled to their own vibration generator 12 via auxiliary drives 15. About a common

Control device 18, the vibration generators 12 are controlled. In addition to the frequency f, phase shifts f among one another are specified for the vibrations generated. As a rule, an opposite synchronization of the tamping units is desired in order to achieve the

Machine frame 3 to minimize retroactive vibrations and noise generation.

[30] For controlling a rotary drive 13 in one

It is necessary for the vibration generator 12 to continuously record the frequencies f or the period T. There are also high levels of track construction

Requirements for the robustness of a sensor arrangement. The evaluation of zero crossings is known with the disadvantage that with it

Frequency changes are detected with a delay. Therefore

provided according to the invention to determine the frequency f or the period T by means of autocorrelation. The basis for this is a measured variable X, which has an approximately periodic function curve of the generated

Represents vibration.

For this purpose, a sensor 19 is arranged, which is coupled magnetically, inductively or capacitively to the vibration generator 12. For example, the distance sensor 19 shown in FIG. 5 comprises a sensitive element which measures a distance to an eccentric lateral surface of an eccentric shaft 14 via inductive coupling. This distance, which changes with the rotating eccentric shaft 14, is the measured variable X, the approximately periodic course of which is subsequently evaluated.

Alternatively, in FIG. 6, a magnetic component 20 is one

Vibration generator 12 is arranged, which is set in rotation by means of the rotary drive 13. This creates a rotating magnetic field that is detected by means of a stationary sensor 19. With this variant the field strength, which changes during one revolution, is recorded and evaluated as measured variable X.

[33] FIG. 7 shows exemplary processing of the measured variable X (or Y for a further vibration generator 12). The upper diagram shows a measured variable curve over time t. This is an approximately periodic course function, whereby as a result of external

Influences disturbances can occur. The aim of the following process steps is to determine the period T or the frequency f = 1 / T. For the measured variable X, measured values x (or y, for a further oscillation generator 12) are recorded in the second diagram, the time interval between the measured values x being specified by a cycle time. For this purpose either a clocked interrogation of the sensor 19 takes place or an analog sensor signal is converted into a signal by means of an analog-digital converter

time-discrete series of measured values x, formed (series values with index i).

[34] Advantageously, faulty

Measured values X _f recognized and eliminated. In addition, it makes sense to improve the sensor signal using an IIR filter, for example of the fourth order. A second-order floch pass removes DC components and a second-order low pass attenuates high-frequency interference in the signal.

[35] In a next step, the measured value curve is interpolated in order to obtain an improved database for the formation of the autocorrelation function. For example, an additional value is interpolated between each recorded measured value x (3rd diagram in FIG. 7).

[36] An exemplary course of the function values y, the autocorrelation function, is shown in FIG. The values y, the autocorrelation function result as follows from a summation of

Measured value products x _n x _ni :

For the present invention, it makes sense if a constant number of measured value products is added up for each function value y.

[37] Advantageously, with each cycle specified by the cycle time, an evaluation of the frequency f or the period duration T is carried out by recalculating the autocorrelation function Calculation of a function value from the sum of the measured value products of the previous calculation, the measured value product with the oldest measured value is subtracted and a new measured value product is added to the current measured value:

The corresponding summation is illustrated in FIG. 9. This simple iterative procedure enables the continuous calculation of the autocorrelation function can be carried out almost in real time with limited computing power.

[38] From the course of the shown in the upper diagram of FIG

Function values y become additional in an optimization step

Function values interpolated. The result is in the diagram below

evident. Since the period T to be determined is determined by the position of the first maximum (except at zero), interpolation is only sensibly carried out in this area. This area to be expected is usually known from the given frequencies in track construction.

[39] To carry out the specified process steps is in a

Evaluation device 21 a computer unit 22 is arranged. In the system illustration in FIG. 10, the computer unit 22 is supplied with the measured values x, from several vibration generators 12 of a controlled working unit 4. On the output side, the individual drives 13 of the

Working unit 4 controlled via assigned power levels 23.

This system structure is shown in more detail in FIG. A first

Assembly 24 comprises computer unit 22, an analog-digital converter 25, a preprocessing unit 26 and communication interfaces 27. Measurement signals from sensors 19 are processed for computer unit 22 via preprocessing unit 26 and analog-digital converter 25. Specifically, time-discrete series of measured values are formed which represent the respective periodic measured value curve.

[41] Via the communication interfaces 27, the computer unit 22 is coupled to a configuration and diagnostic unit 28 and a controller 29 for specifying general control commands. For each Vibration generator 12 is also provided with its own assembly 30, which is coupled to computer unit 22 via communication interfaces 27. Each of these assemblies 30 includes a controller unit 31 and power electronics 32 for controlling the assigned drive 13.

[42] One for four controlled / regulated vibration generators 12

Working unit 4 is an exemplary processing of the measurement signals or measurement values x, shown in FIG. 12. Each vibration generator 12 is a sensor 19 for generating a periodic curve for the

corresponding measured variable X assigned. From this, 34 measured value series are formed with a respective filtering 33 and a sampling rate conversion.

[43] To determine the respective current frequency f, the

Series of measured values of a vibration generator 12 continuously the

Autocorrelation function Y _cc (ΐ) formed. A frequency determination 36 ensues therefrom by means of an extreme value determination 35.

[44] For two oscillation generators 12 in each case, the assigned phase shift f is determined three times in parallel. To this end, a cross-correlation is first formed from the two series of measured values. With the measured values x, of the one vibration generator 12 and the measured values y, of the other vibration generator 12, the following results

Cross correlation function:

From the course of the cross-correlation function Y _cn (ΐ) that is continuously formed in each case, the corresponding extreme value determination 35 is used

Phase shift determination 37.

Claims

1. Method for controlling a rotary drive (13) a

Working unit (4) of a track construction machine (1), with a sensor (19) being used to detect a measured variable (X) derived from a rotation of the drive (13) with an approximately periodic course function

Evaluation device (21) a frequency (f) or period (T) of the

The course function is determined and the frequency (f) or period duration (T) is compared with a nominal value for specifying an actuating signal, characterized in that a series of time-discrete measured values (Xi) is formed for the measured variable (X) and that a computer unit ( 22) an autocorrelation of these measured values (c) is carried out to determine the frequency (f) or period duration (T).

2. The method according to claim 1, characterized in that by means of a further sensor (19) a further measured variable (Y) derived from a rotation of a further drive (13) with an approximately periodic

History function is recorded that a further series of time-discrete measured values (y,) is formed for the further measured variable (Y) and that a cross-correlation of the measured values (x, y,) of both measured variables (X, Y) by means of the computer unit (22)

Determination of a phase shift (f) is carried out.

3. The method according to claim 1 or 2, characterized in that a cycle time is specified to form the time-discrete measured values (c) and that the cycle time determines an evaluation period.

4. The method according to any one of claims 1 to 3, characterized in that with each new measured value (x) an iterative calculation of function values (y,) of a correlation function is carried out by adding a constant number of measured value products.

5. The method according to claim 4, characterized in that for a current calculation of a function value (y,) from the sum of Measured value products of the previous calculation, the measured value product with the oldest measured value is subtracted and a new measured value product is added to the current measured value.

6. The method according to any one of claims 1 to 5, characterized in that interpolated measured values are determined prior to a calculation of function values (y 1) of a correlation function.

7. The method according to any one of claims 1 to 6, characterized in that the measured values (x,) are filtered before a calculation of function values (y,) of a correlation function.

8. The method according to any one of claims 1 to 7, characterized in that interpolated function values (y 1) are determined before a maximum of the correlation function is determined.

9. The method according to any one of claims 1 to 8, characterized in that a distance between a sensitive element of the sensor (19) and an eccentric shaft (14) driven by the drive (13) is detected as the measured variable (X).

10. The method according to any one of claims 1 to 8, characterized in that a field strength acting on the sensor (19) of a magnetic field set in rotation with the drive (13) is detected as the measured variable (X).

11. Device for performing a method according to one of claims 1 to 10, with a sensor (19) for detecting a measured variable (X) derived from a rotation of a drive (13) with an approximately periodic profile function, with an evaluation device (21) for determination a frequency (f) or a period duration (T) of the curve function and with a control device (18) for controlled control of the drive (13), characterized in that time-discrete measured values (x,) of the measured variable (X) are fed to a computer unit (22) are and that in the computer unit (22) an algorithm for performing a Autocorrelation of these measured values (x,) to determine the frequency (f) or

Period duration (T) is set up.

12. The device according to claim 1 1, characterized in that a further sensor (19) is arranged for detecting a further measured variable (Y) derived from a rotation of a further drive (13), that the computer unit (22) discrete-time measured values (y, ) are supplied to the further measured variable (Y) and that an algorithm for performing a cross-correlation of the measured values (x, y,) of both measured variables to determine a phase shift (f) is set up in the computer unit (22).

13. The apparatus of claim 1 1 or 12, characterized in that the computer unit (22) is a microprocessor.

14. Device according to one of claims 1 1 to 13, characterized in that the computer unit (22) with a communication interface (27) is arranged in a first assembly (24) and that a separate assembly (30) for the drive (13) which comprises power electronics (32), a control unit (31) and a communication interface (27), is coupled to the first assembly (24).

15. Device according to one of claims 1 1 to 14, characterized in that the sensor (19) comprises a sensitive element for capacitive or inductive or magnetic detection of the measured variable (X).